CN118014443A - Medicine monitoring method and device - Google Patents
Medicine monitoring method and device Download PDFInfo
- Publication number
- CN118014443A CN118014443A CN202410417620.9A CN202410417620A CN118014443A CN 118014443 A CN118014443 A CN 118014443A CN 202410417620 A CN202410417620 A CN 202410417620A CN 118014443 A CN118014443 A CN 118014443A
- Authority
- CN
- China
- Prior art keywords
- data
- medicine
- raw materials
- crystallinity
- granularity
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 239000003814 drug Substances 0.000 title claims abstract description 178
- 238000000034 method Methods 0.000 title claims abstract description 65
- 229940079593 drug Drugs 0.000 title claims abstract description 50
- 238000012544 monitoring process Methods 0.000 title claims abstract description 39
- 239000002994 raw material Substances 0.000 claims abstract description 159
- 238000004519 manufacturing process Methods 0.000 claims abstract description 56
- 238000012512 characterization method Methods 0.000 claims abstract description 53
- 239000013078 crystal Substances 0.000 claims abstract description 45
- 238000013441 quality evaluation Methods 0.000 claims abstract description 29
- 238000001514 detection method Methods 0.000 claims abstract description 26
- 238000011156 evaluation Methods 0.000 claims abstract description 22
- 238000004140 cleaning Methods 0.000 claims abstract description 10
- 230000000877 morphologic effect Effects 0.000 claims description 29
- 238000004458 analytical method Methods 0.000 claims description 25
- 239000000843 powder Substances 0.000 claims description 20
- 239000011159 matrix material Substances 0.000 claims description 15
- 238000012545 processing Methods 0.000 claims description 14
- 238000003860 storage Methods 0.000 claims description 14
- 239000000825 pharmaceutical preparation Substances 0.000 claims description 13
- 229940127557 pharmaceutical product Drugs 0.000 claims description 12
- 238000009826 distribution Methods 0.000 claims description 11
- 230000008859 change Effects 0.000 claims description 9
- 238000005316 response function Methods 0.000 claims description 8
- 238000004364 calculation method Methods 0.000 claims description 7
- 239000008194 pharmaceutical composition Substances 0.000 claims description 7
- 238000009499 grossing Methods 0.000 claims description 6
- 230000001629 suppression Effects 0.000 claims description 6
- 238000012806 monitoring device Methods 0.000 claims description 5
- 238000005192 partition Methods 0.000 claims description 3
- 235000019580 granularity Nutrition 0.000 description 59
- 239000002245 particle Substances 0.000 description 24
- 238000001303 quality assessment method Methods 0.000 description 16
- 238000012549 training Methods 0.000 description 10
- 238000007637 random forest analysis Methods 0.000 description 9
- 238000004422 calculation algorithm Methods 0.000 description 7
- 230000006870 function Effects 0.000 description 7
- 238000012360 testing method Methods 0.000 description 7
- 238000013145 classification model Methods 0.000 description 6
- 230000008569 process Effects 0.000 description 6
- 239000000047 product Substances 0.000 description 6
- 239000004480 active ingredient Substances 0.000 description 4
- 238000003066 decision tree Methods 0.000 description 4
- 230000000694 effects Effects 0.000 description 4
- 239000000126 substance Substances 0.000 description 4
- 230000002159 abnormal effect Effects 0.000 description 3
- 239000008186 active pharmaceutical agent Substances 0.000 description 3
- 230000008878 coupling Effects 0.000 description 3
- 238000010168 coupling process Methods 0.000 description 3
- 238000005859 coupling reaction Methods 0.000 description 3
- 229940088679 drug related substance Drugs 0.000 description 3
- 230000006872 improvement Effects 0.000 description 3
- 239000000463 material Substances 0.000 description 3
- 238000005457 optimization Methods 0.000 description 3
- 230000000704 physical effect Effects 0.000 description 3
- 238000007781 pre-processing Methods 0.000 description 3
- 238000003908 quality control method Methods 0.000 description 3
- 230000006978 adaptation Effects 0.000 description 2
- 238000005311 autocorrelation function Methods 0.000 description 2
- 238000004891 communication Methods 0.000 description 2
- 238000000354 decomposition reaction Methods 0.000 description 2
- 230000007613 environmental effect Effects 0.000 description 2
- 238000002474 experimental method Methods 0.000 description 2
- 238000005286 illumination Methods 0.000 description 2
- 239000012535 impurity Substances 0.000 description 2
- 230000001788 irregular Effects 0.000 description 2
- 238000009533 lab test Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 239000002547 new drug Substances 0.000 description 2
- 230000006798 recombination Effects 0.000 description 2
- 238000005215 recombination Methods 0.000 description 2
- 238000012216 screening Methods 0.000 description 2
- 230000035945 sensitivity Effects 0.000 description 2
- 238000007619 statistical method Methods 0.000 description 2
- 235000019587 texture Nutrition 0.000 description 2
- 239000002699 waste material Substances 0.000 description 2
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000033228 biological regulation Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 239000003086 colorant Substances 0.000 description 1
- 238000004590 computer program Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 239000006185 dispersion Substances 0.000 description 1
- 238000004090 dissolution Methods 0.000 description 1
- 229940126534 drug product Drugs 0.000 description 1
- 238000001035 drying Methods 0.000 description 1
- 238000003708 edge detection Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 239000012467 final product Substances 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000002844 melting Methods 0.000 description 1
- 230000008018 melting Effects 0.000 description 1
- 238000002156 mixing Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000008447 perception Effects 0.000 description 1
- 230000000737 periodic effect Effects 0.000 description 1
- 238000013139 quantization Methods 0.000 description 1
- 230000009257 reactivity Effects 0.000 description 1
- 230000001105 regulatory effect Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 238000012552 review Methods 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Landscapes
- Investigating Or Analysing Biological Materials (AREA)
Abstract
The invention provides a method and a device for monitoring medicines, and relates to the technical field of medicine monitoring, wherein the method comprises the following steps: acquiring initial data of the medicine, and cleaning the initial data of the medicine to obtain enhanced data of the medicine; identifying crystal characterization factors of the drug enhancement data to obtain the crystallinity of the drug components; calculating the component types of the raw materials according to the component crystallinity of the medicine to obtain the component crystallinity of the raw materials; constructing a quality evaluation model of the medicine raw materials according to the class crystallinity of the raw materials; performing quality evaluation on the medicine raw materials according to the quality evaluation model to obtain an evaluation result; and judging whether the medicine raw materials meet the production requirements according to the evaluation result, producing the medicine raw materials meeting the production requirements, and monitoring the medicines in the production process in real time. The invention can realize quick and accurate evaluation of the quality of the medicine raw materials, and greatly improves the detection efficiency and accuracy.
Description
Technical Field
The invention relates to the technical field of medicine monitoring, in particular to a medicine monitoring method and device.
Background
In the pharmaceutical production process, the quality of the raw materials has a crucial influence on the quality and effect of the final pharmaceutical product. In order to ensure the safety and effectiveness of the medicine, strict quality monitoring of the raw materials is an essential link.
Traditional medicine raw material quality monitoring methods often depend on manual detection and chemical analysis, and the methods are time-consuming and labor-consuming, are easily affected by human factors and environmental factors, so that the accuracy and reliability of detection results cannot be fully ensured.
With the rapid development of artificial intelligence technology, the machine learning algorithm is increasingly widely applied to the field of medicine raw material quality monitoring. By acquiring the data of the raw materials of the medicine and processing and analyzing the data, key information related to the quality of the raw materials can be extracted, and further, the quality of the raw materials can be rapidly and accurately estimated.
However, the existing medicine raw material quality monitoring method still has certain limitations. For example, these methods often only detect certain specific properties of the feedstock and do not allow for a comprehensive assessment of the overall quality of the feedstock.
Disclosure of Invention
In order to more accurately monitor unqualified raw material components in medicines, the invention provides the medicine monitoring method and the medicine monitoring device, which can realize quick and accurate evaluation of medicine raw material quality and greatly improve detection efficiency and accuracy.
In order to solve the technical problems, the technical scheme provided by the invention is as follows:
In a first aspect, a method of monitoring a pharmaceutical product includes the steps of:
acquiring initial data of the medicine, and cleaning the initial data of the medicine to obtain enhanced data of the medicine;
identifying crystal characterization factors of the drug enhancement data to obtain the crystallinity of the drug components;
Calculating the component types of the raw materials according to the component crystallinity of the medicine to obtain the component crystallinity of the raw materials;
constructing a quality evaluation model of the medicine raw materials according to the class crystallinity of the raw materials;
performing quality evaluation on the medicine raw materials according to the quality evaluation model to obtain an evaluation result;
And judging whether the medicine raw materials meet the production requirements according to the evaluation result, producing the medicine raw materials meeting the production requirements, and monitoring the medicines in the production process in real time.
Further, acquiring initial data of the medicine, and performing cleaning treatment on the initial data of the medicine to obtain enhanced data of the medicine, including:
Acquiring powder medicine form data;
carrying out data reconstruction on the morphological data to obtain corresponding morphological reconstruction data;
Acquiring medicine data points in the morphological reconstruction data, and positioning target data to obtain key target data;
and eliminating irrelevant data points in the key target object data to obtain medicine enhancement data.
Further, identifying a crystal characterization factor of the drug enhancement data to obtain a crystallinity of the drug component, comprising:
Calculating the change rate of different characteristic data in the medicine enhancement data to obtain a detection crystal characterization factor;
according to the detected crystal characterization factors, connecting adjacent crystal characterization factors to obtain a morphological data characterization set of the granularity of the raw materials;
acquiring the size and shape information of granularity according to the mutual position information among the data points in the morphological data characterization set so as to obtain an analysis result;
and according to the analysis result, obtaining the crystallinity of the medicine components.
Further, by calculating the change rate of different characteristic data in the drug enhancement data, a detection crystal characterization factor is obtained, which includes:
converting the drug enhancement data into graphic data and calculating a brightness gradient along x and y directions at each point in the graphic data;
setting window size, and carrying out partition calculation on the graphic data according to the windows, wherein for each window W in the graphic data, a matrix M is constructed, wherein, Wherein I x and/>Representing the gradient of the graphics data I in the x and y directions,/>, respectivelyIs the standard deviation of the gaussian function;
from the matrix M for each window, a crystal characterization factor response function R is calculated, wherein, Wherein/>Is a determinant of matrix M,/>Is the trace of matrix M, k is a constant, and the value range of k is 0.04 to 0.06;
And judging and detecting the crystal characterization factor according to the crystal characterization factor response function R and a preset threshold value.
Further, according to the detected crystal characterization factors, connecting adjacent crystal characterization factors to obtain a morphological data characterization set of the granularity of the raw material, including:
According to the detected crystal characterization factor, smoothing the original graph data through a Gaussian filter to obtain smoothed graph data;
According to the smooth graph data, calculating the horizontal gradient of the graph data in the horizontal direction by convolving a Sobel operator in the horizontal direction with the graph data; calculating the vertical gradient of the graphic data in the vertical direction by convolving the Sobel operator in the vertical direction with the graphic data;
for each pixel data point in the graphics data, computing the magnitude and direction of the gradient from the horizontal gradient and the vertical gradient;
According to the amplitude and direction of the gradient, obtaining an edge data set in the graphic data through non-maximum suppression;
according to the edge data set in the graph data, the detected crystal characterization factors are traversed, and according to the positions and distances of the crystal characterization factors in the graph data, the crystal characterization factors are connected by using straight lines or curve segments to form a morphological data characterization set of the granularity of the raw materials.
Further, obtaining size and shape information of granularity according to mutual position information among data points in the morphological data characterization set includes:
Calculating the geometric characteristics of the raw material granularity by calculating the morphological data feature set corresponding to each raw material according to the mutual position information among the data points in the morphological data feature set of the raw material granularity;
and analyzing the shape of the granularity according to the geometric characteristics of the morphological data feature set of each raw material granularity to obtain an analysis result.
Further, according to the analysis result, to obtain the crystallinity of the pharmaceutical composition, the method comprises:
Determining quantized crystallinity indexes, wherein the crystallinity indexes comprise average granularity, average circularity of the shape and standard deviation of granularity distribution;
Comparing and analyzing the obtained granularity and shape data with crystallinity indexes to obtain a comparison result;
the crystallinity of the pharmaceutical composition was evaluated based on the comparison results, and the crystallinity was classified into different grades.
In a second aspect, a monitoring device for a pharmaceutical product comprises:
The acquisition module is used for acquiring the initial data of the medicine and cleaning the initial data of the medicine to obtain the enhanced data of the medicine; identifying crystal characterization factors of the drug enhancement data to obtain the crystallinity of the drug components;
The processing module is used for calculating the component types of the raw materials according to the component crystallinity of the medicine to obtain the component crystallinity of the raw materials; constructing a quality evaluation model of the medicine raw materials according to the class crystallinity of the raw materials; performing quality evaluation on the medicine raw materials according to the quality evaluation model to obtain an evaluation result; and judging whether the medicine raw materials meet the production requirements according to the evaluation result, producing the medicine raw materials meeting the production requirements, and monitoring the medicines in the production process in real time.
In a third aspect, a computing device includes:
One or more processors;
and the storage device is used for storing one or more programs, and when the one or more programs are executed by the one or more processors, the one or more processors realize the monitoring method of the medicine.
In a fourth aspect, a computer readable storage medium has a program stored therein, which when executed by a processor, implements a method of monitoring a drug.
Compared with the prior art, the invention has the beneficial effects that:
According to the scheme, the method can realize quick and accurate evaluation of the quality of the raw materials of the medicines, greatly improve the detection efficiency and accuracy, simultaneously avoid the problems of human errors, environmental interference and the like possibly occurring in the traditional method, and timely discover and solve the quality problem of the raw materials by comprehensively evaluating the quality of the raw materials and monitoring the raw materials in real time in the production process, so that waste and loss caused by the fact that unqualified raw materials enter the production process are avoided, the production cost is reduced, and the safety risk caused by the quality problem of the raw materials is reduced.
Drawings
The accompanying drawings are included to provide a further understanding of the invention, and are incorporated in and constitute a part of this specification.
Fig. 1 is a flow chart of a method for monitoring a drug according to an embodiment of the present invention.
FIG. 2 is a schematic diagram of a monitoring device for a pharmaceutical product provided by an embodiment of the present invention;
Detailed Description
The present invention will be described in further detail with reference to the drawings and examples, in order to make the objects, technical solutions and advantages of the present invention more apparent. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the invention.
In order to better understand the above technical scheme, the following detailed description of the technical scheme of the present invention will be given with reference to the accompanying drawings of the specification and the specific embodiments.
It should be noted that the above steps are only preferred embodiments, and in the specific implementation process, part of the steps may be exchanged without affecting the overall implementation effect, so as to more clearly illustrate the technical solution of the present application, and the following description will explain the present application in a preferred manner.
As shown in fig. 1, the present invention provides a method for monitoring a pharmaceutical product, the method comprising the steps of:
Step 11, acquiring initial data of the medicine, and cleaning the initial data of the medicine to obtain enhanced data of the medicine; specifically, the medicine initial data can be obtained by collecting the graphic data of the medicine raw material, and further, the operation of cleaning the medicine initial data is completed by preprocessing the graphic data of the medicine raw material, so as to obtain medicine enhancement data, and the details of the following embodiments are described by taking the graphic data of the medicine raw material as the medicine initial data;
Step 12, identifying crystal characterization factors of the drug enhancement data to obtain the crystallinity of the drug components; specifically, the angular points in the graph data are used as the crystal characterization factors to obtain the crystallinity of the medicine components;
step 13, calculating the component types of the raw materials according to the component crystallinity of the medicine to obtain the component crystallinity of the raw materials;
Step 14, constructing a quality evaluation model of the medicine raw materials according to the type crystallinity of the raw materials;
step 15, carrying out quality evaluation on the medicine raw materials according to a quality evaluation model to obtain an evaluation result;
and step 16, judging whether the medicine raw materials meet the production requirements according to the evaluation result, producing the medicine raw materials meeting the production requirements, and monitoring the medicines in the production process in real time.
According to the embodiment of the invention, the image data of the medicine raw materials are preprocessed, so that noise and interference factors in the image data can be eliminated, the image data quality is improved, the identification accuracy of the medicine component crystallinity is further improved, and the quality and production compliance of the medicine raw materials can be accurately evaluated. According to the method, the degree of crystallinity of the medicine components is obtained by automatically identifying the corner points of the preprocessing graphic data, so that subjectivity and complexity of traditional manual detection are avoided, the working efficiency is improved, and the possibility of human errors is reduced by improving the degree of automation. The component types of the raw materials are calculated according to the component crystallinity of the medicine, so that more accurate raw material type crystallinity can be obtained, and accurate classification of different types of medicine raw materials is facilitated. By constructing a quality evaluation model based on the crystallinity of the raw materials, the comprehensive quality evaluation can be carried out on the raw materials of the medicine, the quality condition of the raw materials can be reflected more comprehensively, and a reliable basis is provided for production decision. The medicines are monitored in real time in the production process, abnormal conditions in the production process can be found in time, and corresponding measures are taken for adjustment and treatment.
In a preferred embodiment of the present invention, the step 11 may include:
step 111, obtaining powder medicine form data, namely obtaining powder medicine graph data;
Step 112, performing data reconstruction on the morphological data to obtain corresponding morphological reconstruction data, specifically, converting the graphic data into HSV color space to obtain corresponding gray scale graphic data;
Step 113, obtaining medicine data points in the morphological reconstruction data, and positioning target data to obtain key target data, specifically, binarizing gray scale graphic data by using a maximum inter-class variance method to obtain binarized graphic data;
And 114, eliminating irrelevant data points in the key target data to obtain medicine enhancement data, and specifically, obtaining clear-outline image data through outline detection according to the binarized image data.
In the embodiment of the invention, the graphic data of the powder medicine is acquired through the step 111, so that a foundation is provided for subsequent graphic data processing. Step 112 converts the graphic data from the common RGB color space to the HSV color space, and further obtains the corresponding gray scale graphic data, where the HSV color space is more suitable for describing the perception of color by human eyes, and especially when processing the powder medicine graphic data, the brightness information and the color information can be better separated, which is helpful for reducing the influence of the illumination condition change on the graphic data processing result. Step 113 binarizes the gray-scale pattern data using a maximum inter-class variance method (Otsu method). The method can automatically determine a proper threshold value, divides the graphic data into a foreground part and a background part, simplifies the information of the graphic data, highlights the outline and key characteristics of the medicine powder, and is convenient for subsequent analysis and processing of the graphic data. Step 114 processes the binarized graphic data through a contour detection algorithm, so that the contour information of the medicine powder can be accurately extracted, and the graphic data with clear contours can be obtained.
In another preferred embodiment of the present invention, step 111, first, the graphic data of the powder medicine is obtained, which can be accomplished by photographing the powder medicine under controlled illumination conditions using a high-resolution camera, the graphic data should clearly show the texture and color of the powder for the subsequent processing, for example, the powder can be tiled on a smooth surface, ensuring a single background (usually white or black), and then photographed vertically with the camera to obtain high quality graphic data. After the graphic data is acquired, the next step is to convert the graphic data from its original color space (typically RGB) to an HSV color space, which divides the color information into hue (H), saturation (S) and brightness (V), which makes it more straightforward to adjust the brightness or color of the picture, and in this step, the converted brightness value (V) is used as gray scale graphic data. Step 113, performing binarization processing on the gray-scale image data obtained in the previous step by using a maximum inter-class variance method (also called an Otsu's method), specifically including dividing the image data into a foreground and a background by calculating all possible thresholds, and selecting the threshold that maximizes the inter-class variance between the foreground and the background as an optimal threshold, wherein the processed image data only contains two colors of black and white, so that the shape and the edge of the medicine powder are more obvious. Step 114, based on the binarized graphic data, a contour detection algorithm is applied to identify and extract the contour of the powder medicine, and all possible contour lines in the graphic data are found out through an edge detection and graphic algorithm, in this way, a contour map clearly representing the shape and texture of the powder medicine can be obtained. For example, assume that there is a sheet of high-resolution color pattern data of a powder medicine, which is first converted into HSV color space, and a luminance channel is used as gray-scale pattern data. Then, binarizing the gray scale image by Otsu's method to obtain a graphic data for clearly dividing the foreground (medicine powder) and the background. Finally, by means of a contour detection technique, the contour of the pharmaceutical powder can be extracted accurately, for example by calculating the area and circumference of the contour to estimate the particle size distribution of the powder.
In a preferred embodiment of the present invention, the step 12 may include:
Step 121, calculating the change rate of different characteristic data in the medicine enhancement data to obtain a detection crystal characterization factor, specifically, calculating the change rate of a local autocorrelation function of a preprocessing graphic data window to obtain a detection corner point;
Step 122, connecting adjacent crystal characterization factors according to the detected crystal characterization factors to obtain a morphological data feature set of the granularity of the raw material, specifically, connecting adjacent corner points according to the detected corner points by using an edge detector to form an edge contour of the granularity of the raw material;
step 123, obtaining the size and shape information of granularity according to the mutual position information among the data points in the morphological data characterization set to obtain an analysis result, specifically, performing geometric analysis according to the edge profile to evaluate the size and shape of granularity to obtain the analysis result;
And 124, obtaining the crystallinity of the medicine components according to the analysis result.
In the embodiment of the present invention, step 121 detects the corner by calculating the change rate of the local autocorrelation function of the preprocessed graphic data window, and the detection of the corner has higher sensitivity and accuracy, and can accurately locate the corner position in the graphic data. Step 122 uses edge detectors to connect adjacent corner points to form edge contours of raw material granularity, so that complete edge contours of raw material granularity can be effectively extracted, and loss or fracture of edge information is avoided. Step 123 performs geometric analysis according to the edge profile, evaluates the size and shape of the particle size, and can comprehensively describe the geometric characteristics of the particle size of the raw material, including key parameters such as the size and shape. Step 124 obtains the crystallinity of the pharmaceutical composition according to the analysis result. The previous steps provide accurate corner detection, complete edge contour extraction and comprehensive geometric analysis, so that the crystallinity of the medicine components obtained in the previous steps has higher accuracy, and the crystal structure characteristics of the medicine raw materials can be truly reflected.
In a preferred embodiment of the present invention, the step 121 may include:
Step 1211, converting the medicine enhancement data into graphic data, and calculating a luminance gradient in x and y directions at each point in the graphic data, specifically, calculating a luminance gradient in x and y directions at each point in the graphic data;
setting window size, and carrying out partition calculation on the graphic data according to the windows, wherein for each window W in the graphic data, a matrix M is constructed, wherein, Wherein I x and/>Representing the gradient of the graphics data I in the x and y directions,/>, respectivelyIs the standard deviation of the gaussian function;
In step 1212, a corner response function R is calculated from the matrix M for each window, wherein, Wherein/>Is a determinant of matrix M,/>Is the trace of matrix M, k is a constant, and the value range of k is 0.04 to 0.06;
and 1213, judging the detected corner according to the corner response function R and a preset threshold value.
In an embodiment of the present invention, step 1211 calculates the luminance gradient of the graphics data in the x and y directions at each point. Such a calculation can accurately reflect the direction and magnitude of the pixel brightness variation in the graphic data. In step 1211, by constructing the matrix M and introducing a gaussian function to weight, the influence of noise on corner detection can be effectively suppressed. The Gaussian weighting enables the pixels closer to the center of the window to have larger weights, so that the robustness of corner detection is improved. Step 1212 computes a corner response function R from the matrix M for each window. The function can effectively distinguish angular points from non-angular point areas, and the angular point areas have higher response values by calculating determinant and trace of the matrix M and adjusting the determinant and trace by combining a constant k. Step 1213 judges the detected corner according to the corner response function R and a preset threshold, and by setting a suitable threshold, the sensitivity and accuracy of corner detection can be flexibly controlled. When the R value exceeds a preset threshold, the center of the window is considered to have a corner, so that the automatic detection of the corner is realized.
In a preferred embodiment of the present invention, the step 122 may include:
Step 1221, smoothing the original graphic data by a gaussian filter according to the detected crystal characterization factor to obtain smoothed graphic data, specifically, smoothing the original graphic data by a gaussian filter according to the detected corner points to obtain smoothed graphic data;
Step 1222, calculating the horizontal gradient of the graphic data in the horizontal direction by convolving the Sobel operator in the horizontal direction with the graphic data according to the smoothed graphic data; calculating the vertical gradient of the graphic data in the vertical direction by convolving the Sobel operator in the vertical direction with the graphic data;
Step 1223, for each pixel point in the graphics data, calculating the magnitude and direction of the gradient from the horizontal gradient and the vertical gradient;
Step 1224, according to the magnitude and direction of the gradient, obtaining an edge data set in the graphic data through non-maximum suppression, specifically, obtaining an edge line in the graphic data through non-maximum suppression;
step 1225, connecting the corner points by using straight lines or curve segments according to the positions and distances of the corner points in the graphic data by traversing the detected corner points according to the edge lines in the graphic data, so as to form an edge profile of the granularity of the raw materials, namely, a morphological data feature set.
In the embodiment of the present invention, step 1221, firstly, using the detected corner as a reference, smoothing the original graphic data by applying a gaussian filter, and replacing the gray value of each pixel in the graphic data with the weighted average of the gray values of the surrounding pixels by the gaussian filter through convolution operation, thereby removing high-frequency noise and details in the graphic data, obtaining smoothed graphic data, and smoothing the graphic data to reduce the influence of noise on the corner connection. In step 1222, convolution operations are performed on the smooth graphics data in the horizontal direction and the vertical direction respectively using the Sobel operator, so as to obtain gradient values (i.e. the speed and the direction of brightness change) of each pixel point in the horizontal direction and the vertical direction. In step 1223, for each pixel, according to its gradient values in the horizontal direction and the vertical direction, the magnitude of the gradient (i.e. the magnitude of the gradient) is calculated by the pythagorean theorem, and the direction of the gradient (i.e. the angle of the gradient) is calculated by the arctangent function, the magnitude and direction of the gradient being important parameters for determining the edge strength and direction. Step 1224, traversing each pixel in the graphics data, and comparing its gradient magnitude with the gradient magnitude of the adjacent pixel in its gradient direction. If the gradient magnitude of the current pixel point is not the maximum in its gradient direction, it is suppressed (typically set to 0), so that the point where the local gradient is maximum is reserved as an edge point. Non-maxima suppression can refine the edge lines so that the edges are clearer and more accurate. Step 1225, according to the edge lines and the detected corner information obtained in the previous step, connecting the adjacent corner points by using straight line segments or curve segments according to the position and distance relation of the corner points in the graphic data, thereby forming a complete edge profile of the granularity of the raw material, and integrating the discrete corner point information into a continuous edge profile, so as to facilitate the subsequent geometric analysis and crystallinity evaluation.
For example, assume that there is graphic data of a sheet of powder medicine, which contains a plurality of raw material particle sizes. First, corner points of these raw material granularities are detected, via step 121. Then, in step 122, the graphic data is smoothed using a gaussian filter to remove noise; then calculating the horizontal gradient and the vertical gradient of the smooth graph data; further determining the intensity and direction of the edge by calculating the gradient magnitude and direction of each pixel point; applying non-maximum suppression to refine the edge lines; and finally, connecting the detected corner points and the thinned edge lines to form a complete raw material granularity edge profile. Thus, the shape and size of each raw material particle size can be clearly seen.
In a preferred embodiment of the present invention, the step 123 may include:
step 1231, calculating geometric characteristics of each contour according to the edge contour of the raw material granularity, that is, the mutual position information among data points in the morphological data characterization set of the raw material granularity, specifically including: calculating the area and perimeter of the outline through the pixel points on the outline, and determining the coordinates of the pixel points on the outline in the graphic data; calculating the geometric center of the outline according to the coordinates of the pixel points on the outline in the graphic data, and calculating the boundary rectangle and the minimum closed circle of each outline so as to estimate the maximum size and the shape of granularity;
Step 1232, analyzing the shape of the granularity according to the geometric characteristic of each contour to obtain an analysis result, which specifically includes: the regularity and compactness of granularity are evaluated by comparing the actual area of granularity with the area of granularity boundary rectangle.
In the embodiment of the present invention, step 1231, the circumference of the contour is obtained by traversing the pixel points on the raw material granularity edge contour and accumulating the distances between the pixel points. Meanwhile, the number of pixel points in the outline (including the boundary) is counted, and the area of the granularity represented by the outline is obtained by multiplying the number of pixel points by the area of a single pixel point. The coordinates of each pixel point on the contour in the graphic data are recorded, the coordinates are used for subsequent shape analysis and geometric center calculation, the geometric center position of the granularity is calculated by taking an average value according to the coordinates of all the pixel points on the contour, the minimum rectangular area (boundary rectangle) capable of completely containing the granularity contour and the minimum circular area (minimum closed circle) capable of tightly surrounding the granularity contour are determined, and the calculation is helpful for estimating the maximum size and the overall shape of granularity. This step provides quantitative data in many aspects of granularity, location, shape, etc.
In step 1232, the regularity (i.e., the degree to which the shape approximates a standard geometry) and compactness (i.e., the efficiency of utilization of the interior space) of the granularity can be evaluated by comparing the actual area of the granularity with the area of its bounding rectangle. Specifically, if the proportion of the granularity area to the boundary rectangular area is higher, the granularity shape is more regular and compact; otherwise, irregular or loose shapes are indicated. Based on the geometric characteristics of the contour and the above evaluation results, the shape features of the granularity, such as circularity, rectangularity, aspect ratio, etc., are further extracted to more fully describe and analyze the shape of the granularity. The step provides more detailed and accurate shape information for judging the crystallinity of the medicine raw materials through deep analysis of the granularity shape.
For example, assuming that there is graphic data for a batch of powdered drug material, the crystallinity thereof needs to be evaluated by a graphic data processing technique. After the edge profile of the raw material grain size is obtained by the previous step processing, step 123 is entered for geometric characteristic calculation and shape analysis. First, in step 1231, the geometric characteristics of the area, perimeter, geometric center, and bounding rectangle and minimum closed circle of each feedstock particle size are calculated. These data help to understand the size, location and overall shape characteristics of each granularity. Next, in step 1232, the shape of the feedstock particle size is further analyzed using the calculated geometric characteristics. By comparing the ratio of the actual area to the area of the bounding rectangle, it is found that some granularities are more regular and compact in shape, while others exhibit irregular or loose features.
In a preferred embodiment of the present invention, the step 124 may include:
Step 1241, determining a quantified crystallinity index, the crystallinity index comprising an average particle size, an average circularity of the shape, and a standard deviation of the particle size distribution;
Step 1242, comparing the obtained granularity and shape data with crystallinity index to obtain comparison result;
Step 1243, the crystallinity of the pharmaceutical composition is evaluated and classified into different grades according to the comparison result.
In the embodiment of the present invention, in the step 124, the granularity and shape data are compared with the crystal index by clearly quantifying the crystal index, so that the crystal degree of the medicine component can be objectively and accurately evaluated and classified into different grades.
In step 1241, the average size of all the particle sizes of the raw materials is calculated, which can be obtained by taking the average value of the area or diameter of all the particle sizes, the circularity is a measure of the degree to which the particle size shape is nearly circular, which can be obtained by calculating the ratio of the perimeter of each particle size to the perimeter of the circle of equal area, and then taking the average circularity of all the particle sizes, the standard deviation reflecting the degree of dispersion of the particle sizes, i.e., the uniformity of the particle size distribution. This index is obtained by calculating the standard deviation of all particle size data. These quantified crystallinity indices provide an explicit reference standard for subsequent data comparison and crystallinity assessment. Step 1242, comparing the obtained size data of each raw material granularity with the average granularity index, analyzing the distribution condition of the granularity, comparing the circularity data of each granularity with the average circularity index, knowing the integral characteristic of the granularity shape, evaluating the uniformity and consistency of the granularity by comparing the standard deviation index of the granularity distribution, and the comparison analysis of the step can reveal the difference and similarity between the raw material granularity and the ideal crystallinity.
In step 1243, the comparison results in step 1242 are comprehensively analyzed, and the crystallinity of the pharmaceutical composition is classified into different grades according to the comprehensive comparison results, such as "excellent", "good", "general" or "poor", according to the aspects of the granularity, the shape circularity and the granularity distribution, and specific quantization criteria such as the range of the average granularity, the threshold value of the average circularity, and the limit value of the standard deviation can be set for each grade, and through this step, the crystallinity of the pharmaceutical composition can be objectively and accurately assessed, and explicit guidance can be provided for improvement and quality control of the production process.
For example, assuming that there is a batch of pharmaceutical material produced, the crystallinity thereof needs to be assessed. After the previous steps, detailed data on the particle size of the feedstock are obtained. In step 1241, an average particle size of 50 microns, an average circularity of 0.9 (closer to 1 means closer to circular) is determined, and a standard deviation of the particle size distribution is 5 microns. These indices represent the ideal crystallinity criteria. In step 1242, the actual measured particle size and circularity data are compared with the above-mentioned indicators. As a result, it was found that most of the particle sizes were concentrated in the range of 45 to 55. Mu.m, the circularity was generally 0.85 or more, and the standard deviation of the particle size distribution was 4. Mu.m, which was slightly lower than the ideal standard. Finally, in step 1243, the comparison results are combined, and the crystallinity of the batch of drug substance is considered to perform well as a whole, but there is still room for improvement. It is therefore rated "good" and corresponding improvement suggestions are made, such as further optimization of the production process parameters to increase the uniformity of the particle size distribution, etc.
In another preferred embodiment of the present invention, the step 13 may include:
in step 13, the component type of the raw material is calculated from the known component crystallinity of the drug substance, and the raw material type crystallinity is obtained. This process can be divided into several sub-steps:
first, information about various raw material components in a medicine is collected, including their chemical structures, physical properties, their proportions in the medicine, and the like.
The method for determining the relation between the crystallinity of the raw material and the category of the components specifically comprises the following steps: selecting a series of raw material samples with different crystallinity, designing experiments to determine the physical properties (such as solubility, density, melting point, etc.) and chemical properties (such as reactivity, stability, etc.) of the raw materials, determining appropriate component categories (such as high solubility, low solubility, high stability, etc.) according to experimental purposes, conducting experiments under strictly controlled conditions, recording experimental data, collecting relevant information about the crystallinity and component categories of the raw materials, which may include literature data, previous research results or industry standards, conducting statistical analysis on the experimental data to find a correlation coefficient between the crystallinity of the raw materials and the component categories, analyzing the calculated correlation coefficient value, wherein the correlation coefficient has a value ranging from-1 to 1, positive values indicate positive correlation, i.e., the component categories also tend to increase (or more strongly exhibit a certain property) when the crystallinity of the raw materials increases, negative values indicate negative correlation, i.e., the component categories tend to decrease (or decrease in property), and values approaching 0 indicate no obvious linear relationship between the two; key factors affecting the component categories are identified and a determination is made as to whether there is a significant trend.
Using a decision tree method, building a classification model according to the collected data to predict the component category of the raw materials, wherein the method specifically comprises the following steps: collecting a data set, ensuring that the data set contains the crystallinity of raw materials and other characteristics which can influence component categories, dividing the data set into a training set and a testing set, wherein the training set is used for constructing a decision tree model, the testing set is used for evaluating the performance of the model, analyzing the characteristics in the data set, and selecting the characteristics highly related to the component categories as splitting attributes of the decision tree; starting from the training set data, selecting the best split feature, and dividing the data into different subsets according to the value of the feature, recursively performing the above steps for each subset until a stopping condition is met (e.g., all samples in the subset belong to the same class, a preset tree depth is reached, the number of samples in the subset is less than a preset threshold, etc.), and on each leaf node, the most-occurring component class is taken as the predicted class of the node.
Calculating the crystallinity of the raw material category through a classification model, which specifically comprises the following steps: inputting the known medicine component crystallinity data into an established classification model, calculating to obtain component categories of each raw material, and calculating a weighted average crystallinity for each category according to the need to represent the overall crystallinity of the raw material; the accuracy of the classification model is verified using a separate dataset. If a deviation or deficiency of the classification model is found, the previous steps are needed to be returned for adjustment and optimization. For example, assuming that a drug product is composed of three raw materials A, B and C, the crystallinity of each raw material is known to be 80%, 60% and 90%, respectively. Through literature review and experimental data, it was found that high crystallinity feedstocks generally have good solubility and bioavailability, while low crystallinity feedstocks are the opposite. Based on this information, a classification model was established, classifying raw materials with crystallinity higher than 80% into a high solubility class, classifying raw materials with crystallinity lower than 65% into a low solubility class, and classifying raw materials between the two into a medium solubility class. By applying this model, raw materials A and C can be classified into high solubility categories, and raw material B into low solubility categories. The weighted average crystallinity for each class can then be calculated to represent the overall crystallinity of that class of feedstock. In this example, the average crystallinity of the high solubility class is (80% +90%)/2=85%, while the average crystallinity of the low solubility class is 60%.
The invention can improve the key properties of solubility, stability, bioavailability and the like of the medicine by knowing the component types and crystallinity of the raw materials and adjusting the formula of the medicine by a pharmacy. Monitoring the crystallinity of the feedstock during the production process can help ensure consistency and quality of the product. If the crystallinity of the raw material of a certain batch is found to be abnormal, measures can be taken in time to avoid the problem of product quality. By optimizing the selection and use of raw materials, production costs can be reduced while maintaining or improving product quality.
In another preferred embodiment of the present invention, the step 14 may include:
The data set is divided into a training set and a test set, typically using 70% -80% of the data as the training set and the remaining data as the test set, the training set being used to build the model and the test set being used to evaluate the performance of the model.
Setting parameters of a random forest model, such as the number and the maximum depth of decision trees, wherein the parameters can be adjusted according to specific problems, and training the random forest model by using training set data; evaluating the random forest model by using the test set data, calculating indexes such as accuracy, precision, recall rate and the like of the random forest model, performing cross verification, and evaluating the stability and generalization capability of the random forest model; analyzing the prediction result of the random forest model, and optimizing and adjusting the random forest model to obtain a constructed quality assessment model; the constructed quality assessment model is integrated into a drug production or quality control system, and when new drug raw material data is input, the quality assessment model can automatically predict the quality assessment result.
For example, suppose a pharmaceutical manufacturing company needs to perform quality assessment of a certain active ingredient in its raw materials. The crystallinity of the active ingredient is one of the key factors affecting its efficacy and stability. The enterprise collects a collection of historical data including a plurality of quality indicators of raw material crystallinity, purity, solubility, etc., and corresponding quality assessment labels (pass or fail). Through the above step 14, the enterprise builds a drug substance quality assessment model based on the crystallinity of the substance class. In the model building process, a strong correlation between crystallinity and purity and solubility was found. Therefore, in the feature engineering, in addition to directly using crystallinity as a feature, a combination feature such as a ratio of crystallinity to purity, a product of crystallinity and solubility, and the like is extracted. Finally, the enterprise selects a random forest algorithm as the basis of the quality assessment model. By training the training set data and evaluating the testing set data, the model achieves higher accuracy and recall rate. Enterprises deploy the model into their quality control systems, which can automatically predict their quality assessment results when new raw material data is entered.
According to the invention, through automatic prediction, the time and cost of manual detection are reduced, the quality evaluation efficiency is improved, the model can accurately identify unqualified raw materials, and enterprises can more reasonably arrange production plans and raw material purchasing strategies according to quality evaluation results, so that the production efficiency and resource utilization rate are improved.
In another preferred embodiment of the present invention, the step 15 may include:
New data are input into the established quality assessment model, the quality assessment model automatically predicts and calculates according to the input characteristic data, and the quality assessment result of each raw material sample is output. The result may be a specific value, classification label (e.g., pass/fail), or probability value. The evaluation result output by the quality evaluation model is analyzed and converted into a human-readable format or report, and the evaluation result can be classified according to specific standards, such as high quality, medium quality, low quality and the like. In practical application, the accuracy of the model can be verified by comparing with the results of laboratory detection or other quality evaluation methods, and if a large deviation exists between the model prediction result and the actual situation, the model can be returned to the model construction stage to carry out necessary adjustment and optimization.
For example, a pharmaceutical manufacturing company uses the steps described above to construct a quality assessment model based on a random forest algorithm. After the model is built, the enterprise collects a new batch of raw material data and applies the model for quality assessment. Through the prediction of the model, enterprises find that the crystallinity of one batch of raw materials meets the requirement, but other quality indexes such as purity are lower, so that the overall quality evaluation result is poor. Based on this finding, the business decided to suspend the use of the batch and made further laboratory tests. Laboratory test results prove the prediction accuracy of the model, and the batch of raw materials does have the problem that the purity does not reach the standard. By timely finding and avoiding potential quality risks, enterprises avoid drug quality problems and economic losses that may result from using off-specification raw materials.
Through automatic quality assessment, enterprises can quickly know the quality condition of raw materials, and a basis is provided for production decision. Unqualified raw materials are found and processed in time, and the medicine quality risk caused by the quality problem of the raw materials is reduced. Unqualified raw materials are avoided, the rejection rate, the reworking rate and the customer complaint rate are reduced, and the production cost and the quality cost are reduced.
In another preferred embodiment of the present invention, the step 16 may include:
screening the raw materials of the medicines according to the quality evaluation result obtained in the step 15, comparing the evaluation result with preset production standards or regulatory requirements, judging whether the raw materials reach standards, and enabling the qualified raw materials to enter the next production flow, wherein the unqualified raw materials are removed or further processed; qualified medicine raw materials are fed and produced according to the production process flow, and in the production process, production operation standards are strictly adhered to, so that the safety and effectiveness of medicines are ensured.
Monitoring points are set in key links of production, such as steps of mixing, reaction, drying and the like, laboratory detection is carried out by using on-line monitoring equipment or periodic sampling, quality changes of medicines are tracked in real time, monitoring data comprise but are not limited to active ingredient content, impurity level, physical properties (such as granularity and dissolution speed) and the like, statistical analysis is carried out on the data, any abnormal trend or potential problem is identified, and production parameters are adjusted or corrective measures are taken according to analysis results so as to ensure the quality of the products.
For example, a pharmaceutical manufacturing facility may employ step 16 described above when producing a new drug. After quality assessment, the enterprise found that some index in a batch of material was slightly below the production standard. After further analysis, enterprises decide to perform special treatment on the batch of raw materials so as to improve the quality of the batch of raw materials and ensure that the batch of raw materials meet the production requirements. In the production process, enterprises use on-line monitoring equipment to monitor the content of active ingredients and the impurity level of medicines in real time. By adjusting the production parameters in time, enterprises successfully control the quality of the medicines within a predetermined range.
By strictly screening raw materials and monitoring the production process in real time, enterprises can ensure that the quality of the final product meets the preset standard and regulation requirements. Real-time monitoring is helpful for timely finding problems in the production process, and potential quality risks and potential safety hazards are avoided. Through on-line monitoring and data analysis, the enterprise can control production parameters more accurately, unnecessary reworking and waste are reduced, and accordingly production efficiency is improved.
As shown in fig. 2, the present invention also provides a monitoring device 20 for medicine, including:
an acquisition module 21, configured to acquire initial data of a drug, and perform cleaning processing on the initial data of the drug to obtain enhanced data of the drug; identifying crystal characterization factors of the drug enhancement data to obtain the crystallinity of the drug components;
A processing module 22, configured to calculate a component class of the raw material according to the component crystallinity of the medicine, so as to obtain a component class crystallinity of the raw material; constructing a quality evaluation model of the medicine raw materials according to the class crystallinity of the raw materials; performing quality evaluation on the medicine raw materials according to the quality evaluation model to obtain an evaluation result; and judging whether the medicine raw materials meet the production requirements according to the evaluation result, producing the medicine raw materials meeting the production requirements, and monitoring the medicines in the production process in real time.
Embodiments of the present invention also provide a computing device comprising: a processor, a memory storing a computer program which, when executed by the processor, performs the method as described above. All the implementation manners in the method embodiment are applicable to the embodiment, and the same technical effect can be achieved.
Embodiments of the present invention also provide a computer-readable storage medium storing instructions that, when executed on a computer, cause the computer to perform a method as described above. All the implementation manners in the method embodiment are applicable to the embodiment, and the same technical effect can be achieved.
Those of ordinary skill in the art will appreciate that the various illustrative elements and algorithm steps described in connection with the embodiments disclosed herein may be implemented as electronic hardware, or combinations of computer software and electronic hardware. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the solution. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present invention.
It will be clear to those skilled in the art that, for convenience and brevity of description, specific working procedures of the above-described systems, apparatuses and units may refer to corresponding procedures in the foregoing method embodiments, and are not repeated herein.
In the embodiments provided in the present invention, it should be understood that the disclosed apparatus and method may be implemented in other manners. For example, the apparatus embodiments described above are merely illustrative, e.g., the division of the units is merely a logical function division, and there may be additional divisions when actually implemented, e.g., multiple units or components may be combined or integrated into another system, or some features may be omitted or not performed. Alternatively, the coupling or direct coupling or communication connection shown or discussed with each other may be an indirect coupling or communication connection via some interfaces, devices or units, which may be in electrical, mechanical or other forms.
The units described as separate units may or may not be physically separate, and units shown as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.
In addition, each functional unit in the embodiments of the present invention may be integrated in one processing unit, or each unit may exist alone physically, or two or more units may be integrated in one unit.
The functions, if implemented in the form of software functional units and sold or used as a stand-alone product, may be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention may be embodied essentially or in a part contributing to the prior art or in a part of the technical solution, in the form of a software product stored in a storage medium, comprising several instructions for causing a computer device (which may be a personal computer, a server, a network device, etc.) to perform all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a usb disk, a removable hard disk, a ROM, a RAM, a magnetic disk, or an optical disk, etc.
Furthermore, it should be noted that in the apparatus and method of the present invention, it is apparent that the components or steps may be disassembled and/or assembled. Such decomposition and/or recombination should be considered as equivalent aspects of the present invention. Also, the steps of performing the series of processes described above may naturally be performed in chronological order in the order of description, but are not necessarily performed in chronological order, and some steps may be performed in parallel or independently of each other. It will be appreciated by those of ordinary skill in the art that all or any of the steps or components of the methods and apparatus of the present invention may be implemented in hardware, firmware, software, or any combination thereof in any computing device (including processors, storage media, etc.) or network of computing devices, as would be apparent to one of ordinary skill in the art upon reading the present specification.
The object of the invention can thus also be achieved by running a program or a set of programs on any computing device. The computing device may be a well-known general purpose device. The object of the invention can thus also be achieved by merely providing a program product containing program code for implementing said method or apparatus. That is, such a program product also constitutes the present invention, and a storage medium storing such a program product also constitutes the present invention. It is apparent that the storage medium may be any known storage medium or any storage medium developed in the future. It should also be noted that in the apparatus and method of the present invention, it is apparent that the components or steps may be disassembled and/or assembled. Such decomposition and/or recombination should be considered as equivalent aspects of the present invention. The steps of executing the series of processes may naturally be executed in chronological order in the order described, but are not necessarily executed in chronological order. Some steps may be performed in parallel or independently of each other.
While the foregoing is directed to the preferred embodiments of the present invention, it will be appreciated by those skilled in the art that various modifications and adaptations can be made without departing from the principles of the present invention, and such modifications and adaptations are intended to be comprehended within the scope of the present invention.
Claims (10)
1. A method for monitoring a pharmaceutical product, characterized by: the method comprises the following steps:
acquiring initial data of the medicine, and cleaning the initial data of the medicine to obtain enhanced data of the medicine;
identifying crystal characterization factors of the drug enhancement data to obtain the crystallinity of the drug components;
Calculating the component types of the raw materials according to the component crystallinity of the medicine to obtain the component crystallinity of the raw materials;
constructing a quality evaluation model of the medicine raw materials according to the class crystallinity of the raw materials;
performing quality evaluation on the medicine raw materials according to the quality evaluation model to obtain an evaluation result;
And judging whether the medicine raw materials meet the production requirements according to the evaluation result, producing the medicine raw materials meeting the production requirements, and monitoring the medicines in the production process in real time.
2. The method of monitoring a pharmaceutical product according to claim 1, wherein: acquiring initial drug data, and cleaning the initial drug data to obtain enhanced drug data, wherein the method comprises the following steps:
Acquiring powder medicine form data;
carrying out data reconstruction on the morphological data to obtain corresponding morphological reconstruction data;
Acquiring medicine data points in the morphological reconstruction data, and positioning target data to obtain key target data;
and eliminating irrelevant data points in the key target object data to obtain medicine enhancement data.
3. The method of monitoring a pharmaceutical product according to claim 2, wherein: identifying a crystal characterization factor of the drug enhancement data to obtain a crystallinity of the drug component, comprising:
Calculating the change rate of different characteristic data in the medicine enhancement data to obtain a detection crystal characterization factor;
according to the detected crystal characterization factors, connecting adjacent crystal characterization factors to obtain a morphological data characterization set of the granularity of the raw materials;
acquiring the size and shape information of granularity according to the mutual position information among the data points in the morphological data characterization set so as to obtain an analysis result;
and according to the analysis result, obtaining the crystallinity of the medicine components.
4. A method of monitoring a pharmaceutical product according to claim 3, wherein: calculating the change rate of different characteristic data in the medicine enhancement data to obtain detection crystal characterization factors, wherein the detection crystal characterization factors comprise:
converting the drug enhancement data into graphic data and calculating a brightness gradient along x and y directions at each point in the graphic data;
setting window size, and carrying out partition calculation on the graphic data according to the windows, wherein for each window W in the graphic data, a matrix M is constructed, wherein, Wherein I x and/>Representing the gradient of the graphics data I in the x and y directions,/>, respectivelyIs the standard deviation of the gaussian function;
from the matrix M for each window, a crystal characterization factor response function R is calculated, wherein, Wherein/>Is a determinant of matrix M,/>Is the trace of matrix M, k is a constant, and the value range of k is 0.04 to 0.06;
And judging and detecting the crystal characterization factor according to the crystal characterization factor response function R and a preset threshold value.
5. The method of monitoring a pharmaceutical product according to claim 4, wherein: according to the detected crystal characterization factors, connecting adjacent crystal characterization factors to obtain a morphological data characterization set of the granularity of the raw materials, wherein the method comprises the following steps:
According to the detected crystal characterization factor, smoothing the original graph data through a Gaussian filter to obtain smoothed graph data;
According to the smooth graph data, calculating the horizontal gradient of the graph data in the horizontal direction by convolving a Sobel operator in the horizontal direction with the graph data; calculating the vertical gradient of the graphic data in the vertical direction by convolving the Sobel operator in the vertical direction with the graphic data;
for each pixel data point in the graphics data, computing the magnitude and direction of the gradient from the horizontal gradient and the vertical gradient;
According to the amplitude and direction of the gradient, obtaining an edge data set in the graphic data through non-maximum suppression;
according to the edge data set in the graph data, the detected crystal characterization factors are traversed, and according to the positions and distances of the crystal characterization factors in the graph data, the crystal characterization factors are connected by using straight lines or curve segments to form a morphological data characterization set of the granularity of the raw materials.
6. The method of monitoring a pharmaceutical product according to claim 5, wherein: acquiring size and shape information of granularity according to mutual position information among data points in the morphological data characterization set, wherein the method comprises the following steps:
Calculating the geometric characteristics of the raw material granularity by calculating the morphological data feature set corresponding to each raw material according to the mutual position information among the data points in the morphological data feature set of the raw material granularity;
and analyzing the shape of the granularity according to the geometric characteristics of the morphological data feature set of each raw material granularity to obtain an analysis result.
7. The method of monitoring a pharmaceutical product according to claim 6, wherein: according to the analysis result, the crystallinity of the medicine component is obtained, and the method comprises the following steps:
Determining quantized crystallinity indexes, wherein the crystallinity indexes comprise average granularity, average circularity of the shape and standard deviation of granularity distribution;
Comparing and analyzing the obtained granularity and shape data with crystallinity indexes to obtain a comparison result;
the crystallinity of the pharmaceutical composition was evaluated based on the comparison results, and the crystallinity was classified into different grades.
8. A monitoring device of medicine, its characterized in that: comprising the following steps:
The acquisition module is used for acquiring the initial data of the medicine and cleaning the initial data of the medicine to obtain the enhanced data of the medicine; identifying crystal characterization factors of the drug enhancement data to obtain the crystallinity of the drug components;
The processing module is used for calculating the component types of the raw materials according to the component crystallinity of the medicine to obtain the component crystallinity of the raw materials; constructing a quality evaluation model of the medicine raw materials according to the class crystallinity of the raw materials; performing quality evaluation on the medicine raw materials according to the quality evaluation model to obtain an evaluation result; and judging whether the medicine raw materials meet the production requirements according to the evaluation result, producing the medicine raw materials meeting the production requirements, and monitoring the medicines in the production process in real time.
9. A computing device, comprising:
One or more processors;
storage means for storing one or more programs which when executed by the one or more processors cause the one or more processors to implement the method of any of claims 1 to 7.
10. A computer readable storage medium, characterized in that the computer readable storage medium has stored therein a program which, when executed by a processor, implements the method according to any of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410417620.9A CN118014443A (en) | 2024-04-09 | 2024-04-09 | Medicine monitoring method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410417620.9A CN118014443A (en) | 2024-04-09 | 2024-04-09 | Medicine monitoring method and device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN118014443A true CN118014443A (en) | 2024-05-10 |
Family
ID=90948841
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410417620.9A Pending CN118014443A (en) | 2024-04-09 | 2024-04-09 | Medicine monitoring method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN118014443A (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107705261A (en) * | 2017-10-09 | 2018-02-16 | 沈阳东软医疗系统有限公司 | A kind of image rebuilding method and device |
| CN108830849A (en) * | 2018-06-28 | 2018-11-16 | 东北大学 | A kind of rotten stage division of mistake/hypoeutectic Al-Si alloy based on image processing techniques |
| CN108898147A (en) * | 2018-06-27 | 2018-11-27 | 清华大学 | A kind of two dimensional image edge straightened method, apparatus based on Corner Detection |
| CN109034217A (en) * | 2018-07-10 | 2018-12-18 | 成都先进金属材料产业技术研究院有限公司 | Grain size intelligence ranking method based on image recognition depth learning technology |
| CN111699380A (en) * | 2017-12-11 | 2020-09-22 | 法国电力公司 | Method, apparatus and program for processing diffraction image of crystalline material |
| CN115760837A (en) * | 2022-12-06 | 2023-03-07 | 泉州装备制造研究所 | Crystal quality evaluation method and system based on deep neural network |
-
2024
- 2024-04-09 CN CN202410417620.9A patent/CN118014443A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107705261A (en) * | 2017-10-09 | 2018-02-16 | 沈阳东软医疗系统有限公司 | A kind of image rebuilding method and device |
| CN111699380A (en) * | 2017-12-11 | 2020-09-22 | 法国电力公司 | Method, apparatus and program for processing diffraction image of crystalline material |
| CN108898147A (en) * | 2018-06-27 | 2018-11-27 | 清华大学 | A kind of two dimensional image edge straightened method, apparatus based on Corner Detection |
| CN108830849A (en) * | 2018-06-28 | 2018-11-16 | 东北大学 | A kind of rotten stage division of mistake/hypoeutectic Al-Si alloy based on image processing techniques |
| CN109034217A (en) * | 2018-07-10 | 2018-12-18 | 成都先进金属材料产业技术研究院有限公司 | Grain size intelligence ranking method based on image recognition depth learning technology |
| CN115760837A (en) * | 2022-12-06 | 2023-03-07 | 泉州装备制造研究所 | Crystal quality evaluation method and system based on deep neural network |
Non-Patent Citations (1)
| Title |
|---|
| 张辉等: "医药质量检测关键技术及其应用综述", 《仪器仪表学报》, vol. 41, no. 3, 31 March 2020 (2020-03-31) * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| WO2021000524A1 (en) | Hole protection cap detection method and apparatus, computer device and storage medium | |
| US10671833B2 (en) | Analyzing digital holographic microscopy data for hematology applications | |
| CN114549522B (en) | Textile quality detection method based on target detection | |
| CN110097034B (en) | Intelligent face health degree identification and evaluation method | |
| JP6222584B2 (en) | Cell-based tissue analysis | |
| Bejnordi et al. | Stain specific standardization of whole-slide histopathological images | |
| CN108154168B (en) | Comprehensive cargo inspection system and method | |
| US8340372B2 (en) | Image analysis | |
| CN103034838B (en) | A kind of special vehicle instrument type identification based on characteristics of image and scaling method | |
| CN111507426B (en) | Non-reference image quality grading evaluation method and device based on visual fusion characteristics | |
| CN109872335A (en) | A kind of automatic read tablet method and its system for PD-L1 antibody stained slice | |
| CN113658174B (en) | Microkernel histology image detection method based on deep learning and image processing algorithm | |
| CN112613454A (en) | Electric power infrastructure construction site violation identification method and system | |
| JPH08504522A (en) | Method and apparatus for identifying objects using a regular sequence of boundary pixel parameters | |
| CN116228764B (en) | Neonate disease screening blood sheet acquisition quality detection method and system | |
| Sivaraj et al. | Detecting anomalous growth of skin lesion using threshold-based segmentation algorithm and Fuzzy K-Nearest Neighbor classifier | |
| CN113362276A (en) | Visual detection method and system for plate | |
| KR20180045473A (en) | System, method and computer program for melanoma detection using image analysis | |
| Swofford et al. | A method for measuring the quality of friction skin impression evidence: Method development and validation | |
| CN117237747B (en) | Hardware defect classification and identification method based on artificial intelligence | |
| De Gélis et al. | Benchmarking change detection in urban 3D point clouds | |
| CN116309561A (en) | Surface defect identification method based on anti-creeping insulating material | |
| CN118014443A (en) | Medicine monitoring method and device | |
| CN115239947A (en) | Wheat stripe rust severity evaluation method and device based on unsupervised learning | |
| CN113506266A (en) | Method, device and equipment for detecting tongue greasy coating and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination |